chapter five mri sequences 5 mri sequences · table 5- 2: advantages and disadvantages of spin echo...

CHAPTER FIVE MRI SEQUENCES

34

5 MRI SEQUENCES

An MRI sequence is an ordered combination of RF and gradient pulses designed to acquirethe data to form the image. In this chapter I will describe the basic gradient echo, spin echoand inversion recovery sequences used in MRI.The data to create an MR image is obtained in a series of steps. First the tissuemagnetisation is excited using an RF pulse in the presence of a slice select gradient.The other two essential elements of the sequence are phase encoding and frequencyencoding/read out, which are required to spatially localise the protons in the other twodimensions. Finally, after the data has been collected, the process is repeated for a seriesof phase encoding steps.The MRI sequence parameters are chosen to best suit the particular clinical application.The parameters affecting soft tissue contrast are described, and advanced sequences suchas STIR, FLAIR, FISP, and FLASH are briefly introduced at the end of the chapter.

1 Gradient Echo Sequence

The gradient echo sequence is the simplest type of MRI sequence. It consists of a series ofexcitation pulses, each separated by a repetition time TR. Data is acquired at somecharacteristic time after the application of the excitation pulses and this is defined as theecho time TE. TE is the time between the mid-point of the excitation pulse and themid-point of the data acquisition as shown in the sequence diagram, figure 5-1 below. Thecontrast in the image will vary with changes to both TR and TE (see chapter 6).In terms of k-space representation, the simultaneous application of the phase encode andread dephase gradients results in translation from the centre of k-space from A to B. This isfollowed by frequency encoding from B to C via the centre of k-space. Each line of data isFT to extract frequency information from the signal and the process is repeated for differentphase encode gradient strengths. Figure 5-1 below shows the principle of a gradient echosequence 1.

Figure 5- 1: (L) Gradient Echo Sequence and (R) k-space representationGradient Echo imaging does not compensate for B0 inhomogeneities. Therefore, there isan increased sensitivity to T2* decay caused by the lack of a 180º refocusing pulse.Gradient Echo sequences have advantages and disadvantages and these are highlighted in


35

table 5-1 below.Advantages DisadvantagesFast imaging Difficult to generate good T2 contrastLow Flip Angle Sensitive to B0 inhomogeneitiesLess RF power Sensitive to susceptibility effects

Table 5- 1: Advantages and Disadvantages of Gradient Echo Imaging

1 Flip Angle and Ernst Angle

In GE sequences, the choice of flip angle (α) is important for achieving T1-weightedimages. GE sequences generally use small flip angles (< 90°) and very short TRs (typically150 ms) 1. Figure 5-2 below shows that the optimal flip angle depends on the T1 value ofthe tissue being imaged. A short T1 results in a larger optimal flip angle. The dotted linerepresents the best contrast-to-noise ratio for marrow, cartilage and bone for a TR of 100ms.

Figure 5- 2: Choice of Flip Angle Determines Optimum Tissue Contrast 2For each value of T1, there is an optimum flip angle that will give the most signal from asequence where repeated RF excitations are made. This is known as the Ernst Angle andis given by:

αErnst = cos-1[exp(-TR/T1)] Equation 1

2 Spin Echo Sequence

The spin echo (SE) sequence is similar to the GE sequence with the exception that there isan additional 180° refocusing pulse present. This 180° pulse is exactly halfway between theexcitation pulse and the echo (see figure 5-3).Following a 90° RF pulse, the magnetisation vector lies in the transverse plane. Due to T2*dephasing, some spins slow down and others speed up. A 180° pulse is then applied to‘flip’ the spin vectors so that the previously slower vectors are effectively precessing aheadof the previously faster ones. After a further time delay (equal to TE/2), a spin echo isformed, see figure 5-3 below 2.


36

Figure 5- 3: Formation of a Spin Echo

In terms of k-space representation of the spin-echo sequence, the application of phaseencoding and read dephase gradients results in movement from the centre of k-space A toposition B. The 180º pulse reverses the k-space position in both phase and frequencydirections, resulting in movement from B to C. This is followed by frequency encoding fromC to D via the centre of k-space (see figure 5-4). Each line of data is Fourier transformed toextract frequency information from the signal and the process is repeated for differentphase encode steps. Figure 5-4 below shows a spin echo sequence 1.

Figure 5- 4: (L) Spin Echo Sequence and (R) k-space representationThe relative advantages and disadvantages of spin echo sequences are shown in table 5-2below.


37

Advantages DisadvantagesHigh SNR Long scan timesTrue T2 weighting Uses more RF power than a GE

sequenceMinimises susceptibility effects

Table 5- 2: Advantages and Disadvantages of Spin Echo Imaging

3 Inversion Recovery Sequence

Inversion recovery is usually a variant of a SE sequence in that it begins with a 180ºinverting pulse. This inverts the longitudinal magnetisation vector through 180º. When theinverting pulse is removed, the magnetisation vector begins to relax back to B0 as shown infigure 5-5 below:

Figure 5- 5: Inversion RecoveryA 90º excitation pulse is then applied after a time from the 180º inverting pulse known asthe TI (time to inversion). The contrast of the resultant image depends primarily on thelength of the TI as well as the TR and TE. The contrast in the image primarily depends onthe magnitude of the longitudinal magnetisation (as in spin echo) following the chosen delaytime TI. This is shown in figure 5-6 below.

Figure 5- 6: Schematic T1 Weighted Inversion Recovery Diagram for Fat and WaterProtons


38

Figure 5-7 shows the complete inversion recovery pulse sequence.

Figure 5- 7: Inversion Recovery Pulse Sequence

1 Uses of Inversion Recovery Sequence

Contrast is based on T1 recovery curves following the 180º inversion pulse. Inversionrecovery is used to produce heavily T1 weighted images to demonstrate anatomy. The 180ºinverting pulse can produce a large contrast difference between fat and water because fullsaturation of the fat or water vectors can be achieved by utilising the appropriate TI.

4 STIR (Short TI inversion Recovery)

STIR is an inversion recovery pulse sequence that uses a TI that corresponds to the time ittakes fat to recover from full inversion to the transverse plane so that there is nolongitudinal magnetisation corresponding to fat. When the 90º excitation pulse is appliedafter the delay time TI, the signal from fat is nullified. STIR is used to achieve suppressionof the fat signal in a T1 weighted image. A TI of 150-175ms achieves fat suppressionalthough this value varies at different field strengths, (140ms for 1.5T scanner). Figure 5-8below shows that a STIR sequence uses a short TI to suppress the signal from fat in a T2weighted image.

Figure 5- 8: No Fat Vector When the 90° is Applied

5 FLAIR (FLuid Attenuated Inversion Recovery)


39

FLAIR is another variation of the inversion recovery sequence. In FLAIR, the signal fromfluid e.g. cerebrospinal fluid (CSF) is nulled by selecting a TI corresponding to the time ofrecovery of CSF from 180º inversion to the transverse plane. The signal from CSF isnullified and FLAIR is used to suppress the high CSF signal in T2 and proton densityweighted images so that pathology adjacent to the CSF is seen more clearly. A TI ofapproximately 2000 ms achieves CSF suppression at 1.5T. Figure 5-8 above shows that aFLAIR sequence uses a long TI (eg: 2000 ms) and a short TR (eg: 10 ms) to suppress thesignal from water.

6 Fast (Turbo) Spin Echo

Table 5-3 below shows the terms used for Fast Spin Echo sequences by differentmanufacturers.

Manufacturer NameGE Fast Spin Echo (FSE)Siemens, Philips Turbo Spin Echo (TSE)

Table 5- 3: Terms used for Fast Spin Echo by ManufacturerWith each TR in a conventional spin echo, we have a single phase-encoding step. Each ofthe echoes following each 180º pulse is obtained after a single application of thephase-encoding gradient in conventional spin echo. Each echo has its own k-space, andeach time we get an echo, we fill in one line of k-space as shown in figure 5-9 below:

Figure 5- 9: Spin Echo Sequence with 3 Echoes

Fast Spin Echo is a good way of manipulating conventional spin echo to save time.Consider a train of three echoes (Echo Train Length = 3) and only one k-space. Thisk-space will be filled three lines at a time. Instead of having three separate k-spaces, onefor each echo, we will have one k-space using the data from all three echoes as shown inFigure 5-10 below. In conventional spin echo, it took one TR for each line of k-space.Therefore in conventional spin echo, we have to repeat the TR 256 times (for a 256 x 256matrix). We can therefore cut the time by a factor of three with fast spin echo.


40

Figure 5- 10: Diagram above shows that in FSE, k-space is filled three lines at a time in oneTR, two TR and 3TR. After three TRs, 9 lines of k-space will be filled.

For each TR, we will then fill in another three lines into the single k-space. Because thereare a total of 256 lines in k-space, and because during each TR we are filling three lines ofk-space, then we only have to repeat the process 86 times, (256 / 3) to fill 256 lines ofk-space.In conventional spin echo, it took one TR for each line of k-space. Therefore in conventionalspin echo, we have to repeat the TR 256 times. We can therefore cut the time by a factor ofthree (for example) with fast spin echo.

1 Advantages of FSE

With FSE, the scan time is decreased (due to faster scanning) and the SNR is maintainedbecause there are still 256 phase-encoding steps. Motion artefacts will be less severe andthis technique is better able to cope with poorly shimmed magnetic fields than conventionalspin echo.

7 Soft Tissue Contrast in MRI

Contrast is the means by which it is possible to distinguish among soft tissue types owing todifferences in observed MRI signal intensities. For example, in musculoskeletal imaging,there are differences among cartilage, bone and synovial fluid. In neuroimaging, there aredifferences between white and grey matter. The fundamental parameters that affect tissuecontrast are the T1 and T2 values, proton density, tissue susceptibility and dynamics.Tissue pathology will also affect contrast, as will the static field strength, the type ofsequences used, contrast media and the sequence parameters (TR, TE, TI, FA, SNRetc…).

1 T1 Weighting

To demonstrate T1, proton density or T2 contrast, specific values of TR and TE areselected for a given pulse sequence. The selection of appropriate TR and TE weights animage so that one contrast mechanism predominates over the other two.A T1 weighted image is one where the contrast depends predominantly on the differencesin the T1 times between tissues e.g. fat and water. Because the TR controls how far eachvector can recover before it is excited by the next RF pulse, to achieve T1 weighting the TRmust be short enough so that neither fat nor water has sufficient time to fully return to B0. If


41

the TR is too long, both fat and water return to B0 and recover their longitudinalmagnetisation fully. When this occurs, T1 relaxation is complete in both tissues and thedifferences in their T1 times are not demonstrated on the image. The T1 differencesbetween fat and water are shown in figure 5-11 below.

Figure 5- 11: T1 Differences Between Fat and Water

2 T2 Weighting

A T2 weighted image is one where the contrast predominantly depends on the differencesin the T2 times between tissues e.g. fat and water. The TE controls the amount of T2 decaythat is allowed to occur before the signal is received. To achieve T2 weighting, the TE mustbe long enough to give both fat and water time to decay. If the TE is too short, neither fatnor water has had time to decay and therefore the differences in their T2 times are notdemonstrated in the image. The T2 differences between fat and water are shown in figure5-12 below.


42

Figure 5- 12: T2 Differences Between Fat and Water

3 Proton Density Weighting

A proton density image is one where the difference in the numbers of protons per unitvolume in the patient is the main determining factor in forming image contrast. Protondensity weighting is always present to some extent. In order to achieve proton densityweighting, the effects of T1 and T2 contrast must be diminished, so that proton densityweighting can dominate. A long TR allows tissues e.g. fat and water to fully recover theirlongitudinal magnetisation and therefore diminishes T1 weighting. A short TE does not givefat or water time to decay and therefore diminishes T2 weighting. Figure 5-13 below showsa comparison of T1, T2 and PD weighting.

T1PDT2Figure 5- 13: T1, PD and T2 Weighted Axial Brain Images

Table 5-4 below compares the TR and TE for T1, T2 and PD weighted sequences.

Weighting TR TET1 Short ShortT2 Long LongPD Long Short


43

Table 5- 4: Summary of T1, T2 and PD Weighting

4 Contrast to Noise Ratio (CNR)

While improving contrast will generally improve the observer’s ability to differentiate tissues,another factor must be considered. When noise levels are substantial compared to thecontrast, perceived contrast in reduced. Two images may have identical contrast butdifferent noise levels. The image with the lower CNR will be perceived to have a lowercontrast. CNR is defined for tissues A and B as:

CNRAB = (SA – SB) / Noise Equation 2

Where SA and SB are signal intensities for tissues A and B.

8 Advanced Gradient Echo Sequences

Table 5-5 below lists some advanced GE sequences. There are three main types of GEsequences. These are listed below along with the names assigned by two differentmanufacturers.

[1] Free Induction Decay (FID) only (T1 weighted) spoiled Gradient Echo. Sequences of thistype are called Fast Low Angle Shot (FLASH) or Spoiled Gradient Recalled Echo (SPGR).These are referred to as ‘Spoiled GE’ in diagram 5-14 below.[2] FID and Echo (T1/T2) rewound Gradient Echo. Sequences of this type are calledGradient Recalled Acquisition in the Steady State (GRASS) or Fast Imaging with SteadyState Precession (FISP). These are referred to as ‘Rewound GE in diagram 5-14.[3] Echo only (T2) Steady State. Sequences of this type are called ‘Steady State FreePrecession (SSFP)’ or PSIF (FISP reversed).

Sequence Type GE Systems SIEMENSSpoiled GE SPGR FLASHRewound GE GRASS FISPEcho Only SSFP PSIF

Table 5- 5: Advanced Gradient Echo Sequences

Figure 5- 14: GE: Coherent and Incoherent Signal Formation in the Steady State Resultingfrom


44

a Rapid and Regular Train of RF Excitation Pulses 11 FLASH or SPGR

The word “spoiling” refers to the elimination of the steady-state transverse magnetisation.There are various ways of doing this, such as by applying RF spoiling, applying variablegradient spoilers and by lengthening TR. By eliminating the steady state component, onlythe longitudinal component affects the signal in the FLASH technique. This technique lendsitself to reduced T2* weighting and increased T1 weighting. This is true provided that α isalso large. When α is small, the T1 recovery curves play a minor role and proton density(PD) weighting is increased.

2 FISP (Fast Imaging with Steady-state Precession)

In contrast to spin echo (SE), in gradient echo (GRE) there may be residual transversemagnetisation at the end of each cycle remaining for the next cycle. This residualmagnetisation reaches a steady state value after a few cycles. This residual steady statemagnetisation (MSS) is added to the transverse magnetisation created by the next α RFpulse and thus increases the length of the vector in the x-y plane. This then yields more T2*weighting. In other words, tissues with a longer T2 have a longer MSS than do tissues withshorter T2. To preserve this steady state component, a “rewinder gradient” is applied in thephase-encoding direction at the end of the cycle to reverse the effects of thephase-encoding gradient applied at the beginning of the cycle.

3 PSIF (REVERSE Fast Imaging with Steady-state Precession)

This technique yields heavily T2 weighted images. Each α pulse contains some 180º pulseembedded in it that acts like a refocusing pulse. This in turn will result in a spin echo at thetime of the next α pulse. Hence, contrast is determined by T2 (not T2*). Table 5-6 belowshows a comparison of FISP and PSIF sequences.

GE Technique SNR CNR Comments

FISP Highest Best possible T2* Preserves steady-statecomponent

PSIF Lower Provides T2W Gradient recalled SE;TR<TE<2TR

Table 5- 6: Comparison of FISP and PSIF

9 Echo Planar Imaging (EPI)

Single shot EPI requires high performance gradients to allow rapid on and off switching ofthe gradients. The basic idea is to fill k-space in one shot with readout gradient during oneT2* decay or in multiple shots (multishot EPI) by using multiple excitations. Single shot EPIallows oscillating frequency-encoding gradient pulses and complete k-space filling after asingle RF pulse. Increased data processing speeds have allowed EPI to become clinicallywidespread, and it is used in fMRI.

10 3D Gradient Echo for Volumes

3D imaging with contiguous thin slices is feasible by using GE techniques. This type ofimaging is accomplished by addition of a phase-encoding step (NZ) in the slice-selectdirection (z-axis).


45

Total scan time = TR x NY x NEX x NZ

Where NZ is the number of phase encoding steps in the z-direction. NZ is usually a powerof 2 (32, 64, 128 etc). This provides a slab of slices.An extremely short TR makes it possible to do 3D imaging in a reasonable time.For example, the scan time for a 3D GRE technique in the cervical spine with the followingparameters: TR = 30, TE = 13, Angle = 5º, NEX = 1, NY = 256 x 192, NZ = 64.Scan time = (TR)(NEX)(NY)(NZ) = (30)(1)(192)(32) = 184320 ms = 3 min 4 sec.The advantages of 3D GE include rapid volume imaging of thin contiguous slices withoutcrosstalk and increased SNR, because SNR α (NZ)½

11 Fat Suppression

Hydrogen atoms in fat have a lower Larmor frequency than those in water and thisdifference is called the chemical shift. It is approximately 3.5 ppm and is independent of themagnetic field strength 1. In order to improve fat/water soft-tissue contrast it is often usefulto ‘null’ the signal from the fat (‘fat suppression’). One method is to use a STIR (Short TauInversion Recovery) sequence. In this sequence, a 180° pulse is applied to invert thelongitudinal magnetisation, and after a period TI (the Inversion Time), a conventionalgradient-echo sequence is implemented.MRI allows the suppression of the signal coming from fat. This suppression allowsperturbation of tissue contrast to enhance the signal coming from tissues of greater interest.Two types of tissue are commonly suppressed in clinical practice: FAT and WATER.Suppression can be achieved by inversion recovery techniques, chemical saturation orfrequency-selective pre-saturation and spatial pre-saturation in the FoV.

1 Clinical Applications of FATSAT

FATSAT is used to minimise the signal from fat, relative to surrounding tissues. Inmusculoskeletal applications it is used to minimise the fat signal in the marrow emphasisingthe signal from marrow oedema (due to bone contusion, tumour, infection etc.). When usedfor the eye orbits it suppresses the retro-orbital fat to allow detection of an enhancingretro-orbital pathology on a contrast-enhanced study. FATSAT applied to the necksuppresses fat in order to detect and better evaluate the extent of a mass. FATSAT isactually used in many other clinical applications, such as body MRI (liver, kidney imaging)and breast MRI to name a few examples.

12 Summary

This chapter on MRI sequences described the basic gradient echo and spin echosequences as well as inversion recovery. The relation between the flip angle, the Ernstangle and tissue contrast was explained. Other soft tissue contrast factors such as T1, T2and PD weighting were addressed along with the contrast-to-noise ratio. Finally, anoverview of advanced gradient echo sequences was provided. These were divided intothree classes and qualitatively described.

13 References

[1] D. W. McRobbie, E. A. Moore, M. J. Graves, M. R. Prince. MRI – From Picture to Proton(2003). Cambridge University Press.


46

[2] Gandy. S. MRI Physics Lecture Series (2004). Soft Tissue Contrast in MRI. NinewellsHospital, NHS Tayside.

chapter five mri sequences 5 mri sequences · table 5- 2: advantages and disadvantages of spin echo...

Documents